Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
EDISON Data Science Framework (EDSF)
Data Science Professional Education and Training
EDISON Project value proposition and legacy
Yuri Demchenko, EDISON Project
University of Amsterdam
EDISON Initiative, December 2018
EDISON Project (2015-2017)Grant 675419 (INFRASUPP-4-2015: CSA)
Outline
• Background: Data driven research and demand for new skills
– Foundation, recent reports, studies and facts
• EDISON Data Science Framework (EDSF)
– Data Science competences and skills
– Essential Data Scientist professional skills: Thinking and doing like Data Scientist
• Data Science Professional Profiles
• Data Science Body of Knowledge and Model Curriculum
• Use of EDSF and Example curricula
– Competences assessment
– Building Data Science team
• Roadmap recommendations
• References and additional materials
EDISON 2018 Slide DeckData Science Professional Education and
Training2
This work is licensed under the Creative Commons Attribution 4.0 International License.
To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
Visionaries and Drivers:
Seminal works, High level reports, Activities
The Fourth Paradigm: Data-Intensive Scientific Discovery.
By Jim Gray, Microsoft, 2009. Edited by Tony Hey, Kristin Tolle, et al.
http://research.microsoft.com/en-us/collaboration/fourthparadigm/
EDISON 2018 Slide DeckData Science Professional Education and
Training3
Riding the wave: How Europe can gain from
the rising tide of scientific data.
Final report of the High Level Expert Group on
Scientific Data. October 2010.
http://cordis.europa.eu/fp7/ict/e-
infrastructure/docs/hlg-sdi-report.pdf
The Data Harvest: How
sharing research data
can yield knowledge,
jobs and growth.
An RDA Europe Report.
December 2014
https://rd-alliance.org/data-
harvest-report-sharing-data-
knowledge-jobs-and-growth.html
https://www.rd-alliance.org/
HLEG report on European
Open Science Cloud
(October 2016)
https://ec.europa.eu/research/openscienc
e/pdf/realising_the_european_open_scie
nce_cloud_2016.pdf
Emergence of Cognitive Technologies
(IBM Watson, Cortana and others)
Riding the wave (2010): How Europe can gain
from the rising tide of scientific data.
• “Unlocking the full value of scientific data” - Neelie Kroes, Vice-President of the European Commission, responsible for the Digital Agenda
• Just how students will be trained in the future, or how the profession of “data scientist” will be developed, are among the questions the resolution of which is still evolving and will present intellectual challenges for both privately and publicly supported research. - John Wood, HLEG Chair
• Vision 2030: “Our vision is a scientific e-Infrastructure that supports seamless access, use, re-use and trust of data. In a sense, the physical and technical infrastructure becomes invisible and the data themselves become the infrastructure.”
• Proposed set of actions– 4. Train a new generation of data scientists, and broaden public understanding
We urge that the European Commission promote, and the member-states adopt, new policies to foster the development of advanced-degree programmes at our major universities for the emerging field of data scientist. We also urge the member-states to include data management and governance considerations in the curricula of their secondary schools, as part of the IT familiarisation programmes that are becoming common in European education.
EDISON 2018 Slide DeckData Science Professional Education and
Training4
The Data Harvest (2014): How sharing research data
can yield knowledge, jobs and growth
• Planning the data harvest – John Wood
• The era of data driven science
• We want the right minds, with the right data, at the right time. That’s a tall order that requires change in: – The way science works and scientists think
– How scientific institutions operate and interact
– How scientists are trained and employed
Recommendation 2
• DO promote data literacy across society, from researcher to citizen. Embracing these new possibilities requires training and cultural education – inside and outside universities. Data science must be promoted– A first-class science: Data sharing provides the foundation for a new branch of
science.
– Data education: Training in the use, evaluation and responsible management of data needs to be embedded in curricula, across all subjects, from primary school to university.
– Training within EU projects
– Government and public sector training
EDISON 2018 Slide DeckData Science Professional Education and
Training5
HLEG EOSC Report Essentials – Core Data
Experts [ref]
• Core Data Experts is a new class of colleagues with core scientific professional
competencies and the communication skills to fill the gap between the two
cultures.
– Core data experts are neither computer savvy research scientists nor are they hard-
core data or computer scientists or software engineers.
– They should be technical data experts, though proficient enough in the content domain
where they work routinely from the very beginning (experimental design, proposal
writing) until the very end of the data discovery cycle
– Converge two communities:
• Scientists need to be educated to the point where they hire, support and respect Core Data
Experts
• Data Scientists (Core Data Experts) need to bring the value to scientific research and
organisations
• Implementation of the EOSC needs to include instruments to help train, retain
and recognise this expertise,
– In order to support the 1.7 million scientists and over 70 million people working in
innovation.
EDISON 2018 Slide DeckData Science Professional Education and
Training6
[ref] https://ec.europa.eu/research/openscience/pdf/realising_the_european_open_science_cloud_2016.pdf
EOSC Report Recommendations –
Implementation on training and skills
• I2.1: Set initial guiding principles to kick-start the initiative as quickly as
possible. – A first cohort of core data experts should be trained to translate the needs for data driven science
into technical specifications to be discussed with hard-core data scientists and engineers.
– This new class of core data experts will also help translate back to the hard- core scientists the
technical opportunities and limitations
• I3: Fund a concerted effort to develop core data expertise in Europe. – Substantial training initiative in Europe to locate, create, maintain and sustain the required core data
expertise.
– By 2022, to train (hundreds of thousands of) certified core data experts with a demonstrable
effect on ESFRI/e-INFRA activities and prospects for long-term sustainability of this critical
human resource
• Consolidate and further develop assisting material and tools for Data Management Plans and Data
Stewardship plans (including long-term preservation in FAIR status)
• I7: Provide a clear operational timeline to deal with the early preparatory
phase of the EOSC.– Define training needs for the necessary data expertise and draw models for the necessary
training infrastructure
EDISON 2018 Slide DeckData Science Professional Education and
Training7
Initiatives: GO FAIR and IFDS
• Global Open FAIR
– Findable – Accessible – Interoperable - Reusable
• IFDS – Internet of FAIR Data and Services = EOSC
• GO FAIR implementation approach
– GO-TRAIN: Training of data stewards capable of providing FAIR data services
– FAIRdICT: Top Sector Health collaboration with top team ICT
• A critical success factor is availability of expertise in data stewardship
– Training of a new generation of FAIR data experts is urgently needed to provide the necessary capacity
https://www.dtls.nl/fair-data/
https://www.dtls.nl/fair-data/go-fair/
https://www.dtls.nl/fair-data/fair-data-training/
EDISON 2018 Slide DeckData Science Professional Education and
Training8
Industry reports on Data Science Analytics and Data
enabled skills demand
• Final Report on European Data Market Study by IDC (Feb 2017)– The EU data market in 2016 estimated EUR 60 Bln (growth 9.5% from
EUR 54.3 Bln in 2015)
• Estimated EUR 106 Bln in 2020
– Number of data workers 6.1 mln (2016) - increase 2.6% from 2015
• Estimated EUR 10.4 million in 2020
– Average number of data workers per company 9.5 - increase 4.4%
– Gap between demand and supply estimated 769,000 (2020) or 9.8%
• PwC and BHEF report “Investing in America’s data science and analytics talent: The case for action” (April 2017)
– http://www.bhef.com/publications/investing-americas-data-science-and-analytics-talent
– 2.35 mln postings, 23% Data Scientist, 67% DSA enabled jobs
– DSA enabled jobs growing at higher rate than main Data Science jobs
• Burning Glass Technology, IBM, and BHEF report “The Quant Crunch: How the demand for Data Science Skills is disrupting the job Market” (April 2017)
– https://public.dhe.ibm.com/common/ssi/ecm/im/en/iml14576usen/IML14576USEN.PDF
– DSA enabled jobs takes 45-58 days to fill: 5 days longer than average
– Commonly required work experience 3-5 yrs
EDISON 2018 Slide DeckData Science Professional Education and
Training9
Citing EDISON and EDSF
Influenced by EDISON
PwC&BHEF: Demand for DSA enabled jobs
Demand for business people with
analytics skills, not just data
scientists
• Of 2.35 million job postings in
the US
– 23% Data Scientist
– 67% DSA enabled jobs
• Strong demand for managers
and decision makers with Data
Science (data analytics)
skills/understanding
– Challenge to deliver actionable
knowledge and competences to
CEO level managers
EDISON 2018 Slide DeckData Science Professional Education and
Training10
PwC&BHEF: Data Science and Data Analytics
Competences for Managers and Decision Makers
Percent of employers who say data science and analytics skills will be ‘required of all managers’ by 2020
• Source: BHEF and Gallup, Data Science and Analytics Business Survey (December 2016).
EDISON 2018 Slide DeckData Science Professional Education and
Training11
PwC&BHEF: Skills that are tough to find
EDISON 2018 Slide DeckData Science Professional Education and
Training12
To be mapped to Competences, Knowledge, Skills and Personal (soft) Skills
Faster growing jobs require both analytical and social skills
PwC&BHEF: Data Science and Analytics skills, by
2021: The supply-demand challenge
EDISON 2018 Slide DeckData Science Professional Education and
Training13
IBM&BGT: DSA Jobs Time to Fill and Salary
(2016-2017)
• On average, DSA jobs in Professional Services remain open for 53 days, eight days longer than the overall DSA average. (IBM, BGT 2017 Study)
EDISON 2018 Slide DeckData Science Professional Education and
Training14
Closer look at Data related Jobs and Salaries (2016)
Source: The Job Market for Data Professionals, by Robert R Downs, SciDataCon2016
http://www.scidatacon.org/2016/sessions/98/poster/51/
EDISON 2018 Slide DeckData Science Professional Education and
Training15
OECD and UN on Digital Economy and Data Literacy
OECD (Organisation for Economic Coopration and Development)
• Demand for new type of “dynamic self-re-skilling workforce”
• Continuous learning and professional development to become a shared responsibility of workers and organisations
[ref] Skills for a Digital World, OECD, 25-May-2016 http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=DSTI/ICCP/IIS(2015)10/FINAL&docLanguage=En
UN
• Data Revolution Report "A WORLD THAT COUNTS" Presented to Secretary-General (2014)http://www.undatarevolution.org/report/
• Data Literacy is defined as key for digital revolution and Industry 4.0
• Data literacy = critically analyse data collected and data visualised
EDISON 2018 Slide DeckData Science Professional Education and
Training16
PwC study: Millennials at work (2016) - 1
EDISON 2018 Slide DeckData Science Professional Education and
Training17
Confirmed results of previous studies:
• Loyalty-lite to company
– The power of employer brands and the waning
importance of corporate responsibility
• A time of compromise: benefit from individual
package negotiation
• Development and work/life balance are more
important than position or salary
– Work/life balance and diversity promises are
not being kept
• Financial reward is secondary but cash bonuses are valued
• A techno generation avoiding face time and prefer network
communication
• Moving up the ladder faster expectation but often not confirmed
by hard work required
• Generational communication but not without tensions
PwC study: Millennials at work (2016) - 2
• What organisation is an attractive employer?– Opportunities for career progression
– Competitive wages/other financial incentives
– Excellent training/development programmes
• Factors most influenced decision to accept your current job?
– The opportunity for personal development
– The reputation of the organisation
– The role itself
• Which three benefits would you most value from an employer?
– Training and development
– Flexible working hours
– Cash bonuses
EDISON 2018 Slide DeckData Science Professional Education and
Training18
What can employers do?
Business leaders and HR need to work together to:
• Understand this generation
• Get the ‘deal’ right
• Help millennials grow
• Feedback, feedback and more feedback
• Set them free
• Encourage learning
• Allow faster advancement
• Expect millennials to go
Industry 4.0 and demand for new skills
The Fourth Industrial
Revolution, which includes
developments in previously
disjointed fields such as
artificial intelligence and
machine-learning, robotics,
nanotechnology, 3-D printing,
and genetics and
biotechnology, will cause
widespread disruption not
only to business models but
also to labour markets over
the next five years, with
enormous change predicted
in the skill sets needed to
thrive in the new landscape.
This is the finding of a new
report, The Future of Jobs,
published today by the World
Economic Forum.
EDISON 2018 Slide DeckData Science Professional Education and
Training19
Maritime Industry Digital Transformation
and Skills Strategy – Toward Industry 4.0
KNOWLEDGE TRANSFER
LONG-TERM SKILLS STRATEGY
OCEAN LITERACY
CURRICULUM DESIGN
EDUCATION & TRAINING
OCCUPATIONAL PROFILES
Digital Transformation• Digitalisation and IoT• Intelligent Information• Data Management• Digital Assets Manage• Data Driven Optimisation • Agile Continuous
Improvement• Customer Experience• People and skills
Digital Competences/Skills• Automation, robotics,
electrical vehicles• Information and data
literacy • Communication and
collaboration• Digital content creation,
safety • Problem solving and
critical thinking
Big Data
Internet ofThings
SystemIntegration
AdditiveManufacturing
EDISON 2018 Slide DeckData Science Professional Education and
Training20
Challenge for Education:
Sustainable ICT and Data Skills Development
• Educate vs Train– Training is a short term solution
– Education is a basis for sustainable skills development
• Technology focus changes every 3-4 years– Study: 50% of academic curricula are outdated at the time of graduation
• Lack of necessary skills leads to underperforming projects and organisations and loose of competitiveness– Challenge: Policy and decision makers still don’t include planning human factor
(competences and skills) as a part of the technology strategy
• Need to change the whole skills management paradigm– Dynamic (self-) re-skilling: Continuous professional development and shared
responsibility between employer and employee
– Professional and workplace skills and career management as a part of professional orientation
• Millennials factor and changing nature of workforce
EDISON 2018 Slide DeckData Science Professional Education and
Training21
• EDISON Data Science Framework (EDSF)
– Compliant with EU standards on competences and
professional occupations e-CFv3.0, ESCO
– Customisable courses design for targeted education and
training
• Skills development and career management for Core
Data Experts and related data handling professions
• Capacity building and Data Science team design
• Academic programmes and professional training
courses (self) assessment and design
• EU network of Champion universities pioneering Data
Science academic programmes
• Engagement in relevant RDA activities and groups
• Cooperation with International professional organisations
IEEE, ACM, BHEF, APEC (AP Economic Cooperation )
EDISON Products for Data Science Skills
Management and Tailored Education
EDISON 2018 Slide DeckData Science Professional Education and
Training22
EDISON Data Science Framework (EDSF)
EDISON 2018 Slide DeckData Science Professional Education and
Training23
EDISON Framework components– CF-DS – Data Science Competence Framework
– DS-BoK – Data Science Body of Knowledge
– MC-DS – Data Science Model Curriculum
– DSP – Data Science Professional profiles
– Data Science Taxonomies and Scientific Disciplines
Classification
– EOEE - EDISON Online Education Environment
Methodology
• ESDF development based on job market
study, existing practices in academic,
research and industry.
• Review and feedback from the ELG, expert
community, domain experts.
• Input from the champion universities and
community of practice.
What challenges related to skills management the
EDSF can help to address?
1. Guide researchers in using right methods and tools, latest Data
Analytics technologies to extracting value from scientific data
2. Educate and train RI engineers dev to build modern data intensive
research infrastructure and understand trends and project for future
3. Develop new data analytics tools and ensure continuous improvement
(agile model, DevOps)
4. Correctly organise and manage data, make them accessible (adhering
FAIR principles), education new profession of Data Stewards
5. Help managers to facilitate career dev for researchers and organise
effective teams
6. Ensure skills and expertise sustain in organisation
7. Help research institutions to sustain in competition with industry and
business in data science talent hunting
EDISON 2018 Slide DeckData Science Professional Education and
Training24
Competences Map to Knowledge and Skills
• Competence is a demonstrated ability to apply knowledge, skills and attitudes for achieving observable results
EDISON 2018 Slide DeckData Science Professional Education and
Training25
CF-DS/EDSF Approach: Compatibility with e-
CFv3.0 structure and 4-dimensional model
• Competence Framework for Data Science (CF-DS) definition will be built based on European e-Competence framework for IT (e-CFv3.0)
– Linking scientific research lifecycle, organizational roles, competences, skills and knowledge
– Defining Data Science Body of Knowledge (DS-BoK)
– Mapping CF-DS and DS-BoK to academic disciplines
EDISON 2018 Slide DeckData Science Professional Education and
Training26
• Multiple use of e-CFv3.0 within ICT organisations
• Provides basis for individual career path, competence assessment, training and certification
• EDISON CF-DS will be used for defining DS-BoK and MC-DS, linking organizational functions and required knowledge
• Provide basis for individual (self) training and certification
Data Scientist definition
Based on the definitions by NIST SP1500 – 2015, extended by EDISON
• A Data Scientist is a practitioner who has sufficient
knowledge in the overlapping regimes of expertise in
business needs, domain knowledge, analytical skills,
and programming and systems engineering expertise
to manage the end-to-end scientific method process
through each stage in the big data lifecycle till the delivery
of an expected scientific and business value to
organisation or project.
EDISON 2018 Slide DeckData Science Professional Education and
Training27
• Core Data Science competences and skills groups
– Data Science Analytics (including Statistical Analysis, Machine Learning, Business Analytics)
– Data Science Engineering (including Software and Applications Engineering, Data
Warehousing, Big Data Infrastructure and Tools)
– Domain Knowledge and Expertise (Subject/Scientific domain related)
• EDISON identified 2 additional competence groups demanded by organisations
– Data Management, Data Governance, Stewardship, Curation, Preservation
– Research Methods and/vs Business Processes/Operations
• Data Science professional skills: Thinking and acting like Data Scientist – required to
successfully develop as a Data Scientist and work in Data Science teams
Data Science Competence Groups - Research
EDISON 2018 Slide DeckData Science Professional Education and
Training28
Scientific Methods
• Design Experiment
• Collect Data
• Analyse Data
• Identify Patterns
• Hypothesis Explanation
• Test Hypothesis
Business Operations
• Operations Strategy
• Plan
• Design & Deploy
• Monitor & Control
• Improve & Re-design
Data Science Competences
include 5 groups
• Data Science Analytics
• Data Science Engineering
• Domain Knowledge and Expertise
• Data Management
• Research Methods and Project
Management
– Business Process Management (biz)
Scientific Methods
• Design Experiment
• Collect Data
• Analyse Data
• Identify Patterns
• Hypothesise Explanation
• Test Hypothesis
Business Process
Operations/Stages
• Design
• Model/Plan
• Deploy & Execute
• Monitor & Control
• Optimise & Re-design
Data Science Competences Groups – Business
EDISON 2018 Slide DeckData Science Professional Education and
Training29
Data Science Competences
include 5 groups
• Data Science Analytics
• Data Science Engineering
• Domain Knowledge and Expertise
• Data Management
• Research Methods and Project
Management
– Business Process Management (biz)
Identified Data Science Competence GroupsData Science Analytics (DSDA)
Data Science Engineering (DSENG)
Data Management and Governance (DSDM)
Research/Scientific Methods and Project Management (DSRMP)
Data Science Domain Knowledge, e.g. Business Analytics (DSDK/DSBPM)
0 Use appropriate data analytics and statistical techniques on available data to deliver insights into research problem or org. processes and support decision making
Use engineering principles and modern computer technology to research, design, implement new data analytics applications, develop experiments, processes, instruments, systems and infrastructures to support data handling during the whole data lifecycle
Develop and implement data management strategy for data collection, storage, preservation, and availability for further processing.
Create new understandings and capabilities by using the scientific method (hypothesis, test/artefact, evaluation) or similar engineering methods to discover new approaches to create new knowledge and achieve research or organisational goals
DSDK/DSBA Use domain knowledge (scientific or business) to develop relevant data analytics applications; adopt general Data Science methods to domain specific data types and presentations, data and process models, organisational roles and relations
1 DSDA01Effectively use variety of data analytics techniques
DSENG01Use engineering principles (general and software) to research, design, develop and implement new instruments and applications
DSDM01Develop and implement data strategy, in particular, Data Management Plan (DMP)
DSRMP01Create new understandings and capabilities by using scientific/ research methods
DSBPM01Understand business and provide insight, translate unstructured business problems into an abstract mathematical framework
2 DSDA02Apply designated quantitative techniques
DSENG02Develop and apply computer methods to domain related problems
DSDM02Develop data models including metadata
DSRMP02Direct systematic study toward a fuller knowledge or understanding of the observable facts
DSBPM02Participate strategically and tactically in financial decisions
3 DSDA03Pull together data from diff sources …
DSENG03Develop and prototype data analytics applications
DSDM03Collect integrate data
DSRMP03Undertakes creative work
DSBPM03Provides support services to other
4 DSDA04Use diff perform techniques
DSENG04Develop, deploy operate Big Data storage
DSDM04Maintain repository
DSRMP04Translate strategies into actions
DSBPM04Analyse data for marketing
5 DSDA05Develop analytics applic
DSENG05Apply security mechanisms
DSDM05Visualise cmplx data
DSRMP05Contribute to organis goals
DSBPM05Analyse optimise customer relatio
6 DSDA06Visualise results of analysis, dashboards
DSENG06Design, build, operate SQL and NoSQL
DSRM06Develop and manage prolicies
DSRMP06Develop and guide data driven projects
DSBPM06Analyse data for marketing 30EDISON 2018 Slide Deck
Data Science Professional Education and
Training
Identified Data Science Skills/Experience Groups
Skills Type A – Based on knowledge acquired
• Group 1: Skills/experience related to competences
– Data Analytics and Machine Learning
– Data Management/Curation (including both general data management and scientific data management)
– Data Science Engineering (hardware and software) skills
– Scientific/Research Methods or Business Process Management
– Application/subject domain related (research or business)
• Group 2: Mathematics and statistics
– Mathematics and Statistics and others
Skills Type B – Base on practical or workplace experience
• Group 3: Big Data (Data Science) tools and platforms
– Big Data Analytics platforms
– Mathematics & Statistics applications & tools
– Databases (SQL and NoSQL)
– Data Management and Curation platform
– Data and applications visualisation
– Cloud based platforms and tools
• Group 4: Data analytics programming languages and IDE
– General and specialized development platforms for data analysis and statistics
• Group 5: Soft skills and Workplace skills
– Data Science professional skills: Thinking and Acting like Data Scientist
– 21st Century Skills: Personal, inter-personal communication, team work, professional network
EDISON 2018 Slide DeckData Science Professional Education and
Training31
Example Data Science
Competences Definition
Compliant with e-CFv3.0
EDISON 2018 Slide DeckData Science Professional Education and
Training32
Group 5: Soft skills and Workplace skills
• Data Science professional skills: Thinking and Acting like Data Scientist
• 21st Century Skills: Personal, inter-personal communication, team work,
professional network
• Data Scientist and Subject Domain Specialist
EDISON 2018 Slide DeckData Science Professional Education and
Training33
Data Science Professional Skills:
Thinking and Acting like Data Scientist
1. Recognise value of data, work with raw data, exercise good data intuition, use SN and open data
2. Accept (be ready for) iterative development, know when to stop, comfortable with failure, accept the
symmetry of outcome (both positive and negative results are valuable)
3. Good sense of metrics, understand importance of the results validation, never stop looking at
individual examples
4. Ask the right questions
5. Respect domain/subject matter knowledge in the area of data science
6. Data driven problem solver and impact-driven mindset
7. Be aware about power and limitations of the main machine learning and data analytics algorithms
and tools
8. Understand that most of data analytics algorithms are statistics and probability based, so any
answer or solution has some degree of probability and represent an optimal solution for a number of
variables and factors
9. Recognise what things are important and what things are not important (in data modeling)
10. Working in agile environment and coordinate with other roles and team members
11. Work in multi-disciplinary team, ability to communicate with the domain and subject matter experts
12. Embrace online learning, continuously improve your knowledge, use professional networks and
communities
13. Story Telling: Deliver actionable result of your analysis
14. Attitude: Creativity, curiosity (willingness to challenge status quo), commitment in finding new
knowledge and progress to completion
15. Ethics and responsible use of data and insight delivered, awareness of dependability (data scientist is
a feedback loop in data driven companies)
EDISON 2018 Slide DeckData Science Professional Education and
Training34
Data Science Professional Skills:
Thinking and Acting like Data Scientist (1)
1. Recognise value of data, work with raw data, exercise good data intuition, use
SN and Open Data
2. Accept (be ready for) iterative development, know when to stop, comfortable
with failure, accept the symmetry of outcome (both positive and negative results
are valuable)
3. Good sense of metrics, understand importance of the results validation, never
stop looking at individual examples
4. Ask the right questions
5. Respect domain/subject matter knowledge in the area of data science
6. Data driven problem solver and impact-driven mindset
7. Be aware about power and limitations of the main machine learning and data
analytics algorithms and tools
8. Understand that most of data analytics algorithms are statistics and
probability based, so any answer or solution has some degree of probability
and represent an optimal solution for a number of variables and factors
EDISON 2018 Slide DeckData Science Professional Education and
Training35
Data Science Professional Skills:
Thinking and Acting like Data Scientist (2)
1. S
2. S
3. s
4. Ss
5. S
6. S
7. S
8. s
9. Recognise what things are important and what things are not important (in data
modeling)
10. Working in agile environment and coordinate with other roles and team members
11. Work in multi-disciplinary team, ability to communicate with the domain and subject
matter experts
12. Embrace online learning, continuously improve your knowledge, use professional
networks and communities
13. Story Telling: Deliver actionable result of your analysis
14. Attitude: Creativity, curiosity (willingness to challenge status quo), commitment in finding
new knowledge and progress to completion
15. Ethics and responsible use of data and insight delivered, awareness of dependability
(data scientist is a feedback loop in data driven companies)
EDISON 2018 Slide DeckData Science Professional Education and
Training36
21st Century Skills (DARE & BHEF & EDISON)
1. Critical Thinking: Demonstrating the ability to apply critical thinking skills to solve problems and make effective decisions
2. Communication: Understanding and communicating ideas
3. Collaboration: Working with other, appreciation of multicultural difference
4. Creativity and Attitude: Deliver high quality work and focus on final result, intitiative, intellectual risk
5. Planning & Organizing: Planning and prioritizing work to manage time effectively and accomplish assigned tasks
6. Business Fundamentals: Having fundamental knowledge of the organization and the industry
7. Customer Focus: Actively look for ways to identify market demands and meet customer or client needs
8. Working with Tools & Technology: Selecting, using, and maintaining tools and technology to facilitate work activity
9. Dynamic (self-) re-skilling: Continuously monitor individual knowledge and skills as shared responsibility between employer and employee, ability to adopt to changes
10. Professional networking: Involvement and contribution to professional network activities
11. Ethics: Adhere to high ethical and professional norms, responsible use of power data driven technologies, avoid and disregard un-ethical use of technologies and biased data collection and presentation
EDISON 2018 Slide DeckData Science Professional Education and
Training37
Data Scientist and Subject Domain Specialist
• Subject domain components
– Model (and data types)
– Methods
– Processes
– Domain specific data and presentation/visualization methods
– Organisational roles and relations
• Data Scientist is an assistant to Subject Domain Specialists
– Translate subject domain Model, Methods, Processes into abstract data driven form
– Implement computational models in software, build required infrastructure and tools
– Do (computational) analytic work and present it in a form understandable to subject
domain
– Discover new relations originated from data analysis and advice subject domain
specialist
– Present/visualise information in domain related actionable way
– Interact and cooperate with different organizational roles to obtain data and deliver
results and/or actionable data
EDISON 2018 Slide DeckData Science Professional Education and
Training38
Data Science and Subject Domains
EDISON 2018 Slide DeckData Science Professional Education and
Training39
• Models (and data types)
• Methods
• Processes
Domain specific components
Domain specific
data & presentation
(visualization)
Organisational
roles
• Abstract data driven
math&compute models
• Data Analytics
methods
• Data and Applications
Lifecycle Management
Data Science domain components
Data structures &
databases/storage
Visualisation
Cross-
organisational
assistive role
Data Scientist/Data Steward functions is to translate between two domains
Data Scientist role is to maintain the Data Value Chain (domain specific): • Data Integration => Organisation/Process/Business Optimisation => Innovation
Practical Application of the CF-DS
• Basis for the definition of the Data Science Body of Knowledge (DS-BoK) and Data Science Model Curriculum (MC-DS)
– CF-DS => Learning Outcomes (MC-DS) => Knowledge Areas (DS-BoK)
– CF-DS => Data Science taxonomy of scientific subjects and vocabulary
• Data Science professional profiles definition
– Extend existing EU standards and occupations taxonomies: e-CFv3.0, ESCO, others
• Professional competence benchmarking
– For customizable training and career development
– Including CV or organisational profiles matching
• Professional certification
– In combination with DS-BoK professional competences benchmarking
• Vacancy construction tool for job advertisement (for HR)
– Using controlled vocabulary and Data Science Taxonomy
EDISON 2018 Slide DeckData Science Professional Education and
Training40
Data Science Professions Family
EDISON 2018 Slide DeckData Science Professional Education and
Training41
Icons used: Credit to [ref] https://www.datacamp.com/community/tutorials/data-science-industry-infographic
DSP Profiles mapping to ESCO Taxonomy
High Level Groups
• DSP Profiles mapping to corresponding CF-DS Competence Groups – Relevance level from 5 – maximum to 1 – minimum
EDISON 2018 Slide DeckData Science Professional Education and
Training42
CF-DS and Data Science Professional Profiles
EDISON 2018 Slide DeckData Science Professional Education and
Training43
Whole range of
DSA enabled Domain
related Profiles
Data Management
and Stewardship
Example DS Professional Profile Definition
(compliant with CWA)
EDISON 2018 Slide DeckData Science Professional Education and
Training44
Profile title Gives a commonly used name to a profile. TEMPLATE
Summary statement Indicates the main purpose of the profile.
The purpose is to present to stakeholders and users a brief, concise
understanding of the specified ICT Profile. It should be understandable
by ICT professionals, ICT managers and Human Resource personnel.
It should provide a statement of the job’s main activity.
Mission Describes the rationale of the profile.
The purpose is to specify the designated job role defined in the ICT Profile.
Deliverables Accountable (A) Responsible (R) Contributor (C)
Specifies the Profile by key deliverables.
The purpose is to illuminate the ICT Profiles and to explain relevance
including the perspective from a non-ICT point of view.
Main task/s Provides a list of typical tasks to be performed by the profile.
A task is an action taken to achieve a result within a broadly defined
context. Tasks may be associated with deadlines, resources, goals,
specifications and/or the expected results.
e-CF
competences
assigned
Provides a list of necessary competences (from the e-CF) to carry out the
mission.
Must include 1 up to 5 competences.
Level assignment is important. Can be (usually) 1 or (maximum) 2 levels.
KPI Area Based upon KPIs (Key Performance Indicators) KPI area is a more generic
indicator, congruent with the overall profile granularity level. It is deployed
to add depth to the mission.
Not prescriptive. Non-specific measurements. Use general examples.
The principle is to provide KPI areas (which are stable, general and long
lasting) providing users with an inspiration to enable development of
specific KPI’s for specific roles
Must be related to the key deliverables in order to measure them.
EDSF for Education and Training
• Foundation and methodological base
– Data Science Body of Knowledge (DS-BoK)
• Taxonomy and classification of Data Science related scientific subjects
– Data Science Model Curriculum (MC-DS)
• Set Learning Units mapped to CF-DS Learning and DS-BoK Knowledge Areas/Units
– Instructional methodologies and teaching models
• Platforms and environment
– Virtual labs, datasets, developments platforms
– Online education environment and courses management
• Services
– Individual benchmarking and profiling tools (competence assessment)
– Knowledge evaluation tools
– Certifications and training for self-made Data Scientists practitioners
– Education and training marketplace: Courses catalog and repository
EDISON 2018 Slide DeckData Science Professional Education and
Training45
Data Science Body of Knowledge (DS-BoK)
DS-BoK Knowledge Area Groups (KAG)
• KAG1-DSA: Data Analytics group including
Machine Learning, statistical methods,
and Business Analytics
• KAG2-DSE: Data Science Engineering group
including Software and infrastructure engineering
• KAG3-DSDM: Data Management group including data curation,
preservation and data infrastructure
• KAG4-DSRM: Research Methods and Project Management group
• KAG5-DSBA: Business Analytics and Business Intelligence
• KAG* - DSDK: Data Science domain knowledge to be defined by related
expert groups
EDISON 2018 Slide DeckData Science Professional Education and
Training46
Data Science Body of Knowledge (1)
KA Groups Suggested DS Knowledge Areas (KA) Knowledge Areas from existing BoK and CCS2012 scientific subject groups
KAG1-DSDA: Data Science Analytics
KA01.01 (DSDA.01/SMDA) Statistical methods for data analysisKA01.02 (DSDA.02/ML) Machine LearningKA01.03 (DSDA.03/DM) Data MiningKA01.04 (DSDA.04/TDM) Text Data MiningKA01.05 (DSDA.05/PA) Predictive AnalyticsKA01.06 (DSDA.06/MODSIM) Computational modelling, simulation and optimisation
There is no formal BoK defined for Data Analytics.
Data Science Analytics related scientific subjects from CCS2012:CCS2012: Computing methodologiesCCS2012: Mathematics of computingCCS2012: Computing methodologies
KAG2-DSENG: Data Science Engineering
KA02.01 (DSENG.01/BDI) Big Data Infrastructure and TechnologiesKA02.02 (DSENG.02/DSIAPP) Infrastructure and platforms for Data Science applicationsKA02.03 (DSENG.03/CCT) Cloud Computing technologies for Big Data and Data AnalyticsKA02.04 (DSENG.04/SEC) Data and Applications securityKA02.05 (DSENG.05/BDSE) Big Data systems organisation and engineeringKA02.06 (DSENG.06/DSAPPD) Data Science (Big Data) applications designKA02.07 (DSENG.07/IS) Information systems (to support data driven decision making)
ACM CS-BoK selected KAs:AR - Architecture and Organization (including computer architectures and network architectures)CN - Computational ScienceIM - Information ManagementSE - Software Engineering (can be extended with specific SWEBOK KAs)
SWEBOK selected KAs• Software requirements• Software design• Software engineering process• Software engineering models and methods• Software qualityData Science Analytics related scientific subjects
from CCS2012
EDISON 2018 Slide DeckData Science Professional Education and
Training47
Data Science Body of Knowledge (2)
KA Groups Suggested DS Knowledge Areas (KA) Knowledge Areas from existing BoK and CCS2012 scientific subject groups
KAG3-DSDM: Data Management
KA03.01 (DSDM.01/DMORG) General principles and concepts in Data Management and organisationKA03.02 (DSDM.02/DMS) Data management systemsKA03.03 (DSDM.03/EDMI) Data Management and Enterprise data infrastructureKA03.04 (DSDM.04/DGOV) Data GovernanceKA03.05 (DSDM.05/BDST0R) Big Data storage (large scale) KA03.06 (DSDM.05/DLIB) Digital libraries and archives
DM-BoK selected KAs(1) Data Governance, (2) Data Architecture, (3) Data Modelling and Design, (4) Data Storage and Operations, (5) Data Security, (6) Data Integration and Interoperability, (7) Documents and Content, (8) Reference and Master Data, (9) Data Warehousing and Business Intelligence, (10) Metadata, and (11) Data Quality.
KAG4-DSRM: Research Methods and Project Management
KA04.01 (DSRMP.01/RM) Research MethodsKA04.01 (DSRMP.02/PM) Project Management
There are no formally defined BoK for research methodsPMI-BoK selected KAs• Project Integration Management • Project Scope Management • Project Quality • Project Risk Management
KAG5-DSBPM: Business Analytics
KA05.01 (DSBA.01/BAF) Business Analytics FoundationKA05.02 (DSBA.02/BAEM) Business Analytics organisation and enterprise management
BABOK selected KAs *)Business Analysis Planning and MonitoringRequirements Life Cycle ManagementSolution Evaluation and improvements recommendation
EDISON 2018 Slide DeckData Science Professional Education and
Training48
Data Science Model Curriculum (MC-DS)
Data Science Model Curriculum includes
• Learning Outcomes (LO) definition based on CF-DS– LOs are defined for CF-DS competence groups and for all
enumerated competences
– Knowledge levels: Familiarity, Usage, Assessment (based in Bloom’s Taxonomy)
• LOs mapping to Learning Units (LU)– LUs are based on CCS(2012) and universities best practices
– Data Science university programmes and courses inventory (interactive)http://edison-project.eu/university-programs-list
• LU/course relevance: Mandatory Tier 1, Tier 2, Elective, Prerequisite
• Learning methods and learning models (in progress)
EDISON 2018 Slide DeckData Science Professional Education and
Training49
Learning methods and learning models
• Bloom’s Taxonomy and Cognitive learning activities
– BT application areas and limitations
• Constructive Alignment and Intended Learning Outcome (ILO)
– ILO is formulated from the student perspective
– Outcome Based Learning (OBL)
• Other education technologies for teaching in fast technology changing
world
– Project Based Learning (PBL)
– Flipped classroom
– Activating teaching and
activating strategies
EDISON 2018 Slide DeckData Science Professional Education and
Training50
Bloom’s Taxonomy and Knowledge Levels for
MC-DS
EDISON 2018 Slide DeckData Science Professional Education and
Training51
Level Action Verbs
Familiarity Choose, Classify, Collect, Compare, Configure, Contrast, Define, Demonstrate, Describe, Execute, Explain, Find, Identify, Illustrate, Label, List, Match, Name, Omit, Operate, Outline, Recall, Rephrase, Show, Summarize, Tell, Translate
Usage Apply, Analyze, Build, Construct, Develop, Examine, Experiment with, Identify, Infer, Inspect, Model, Motivate, Organize, Select, Simplify, Solve, Survey, Test for, Visualize
Assessment Adapt, Assess, Change, Combine, Compile, Compose, Conclude, Criticize, Create, Decide, Deduct, Defend, Design, Discuss, Determine, Disprove, Evaluate, Imagine, Improve, Influence, Invent, Judge, Justify, Optimize, Plan, Predict, Prioritize, Prove, Rate, Recommend, Solve
Extended Bloom’s Taxonomy
Consolidated presentation
of learning levels, action
verbs, and associated
learning instruments
[ref] Anderson, L. W., &
Krathwohl, D. R. (2001). A
taxonomy for learning,
teaching, and assessing,
Abridged Edition. Boston, MA:
Allyn and Bacon.
EDISON 2018 Slide DeckData Science Professional Education and
Training52
Bloom’s Taxonomy – Cognitive Activities
KnowledgeExhibit memory of previously learned materials by recalling facts, terms, basic concepts and answers
• Knowledge of specifics - terminology, specific facts
• Knowledge of ways and means of dealing with specifics - conventions, trends and sequences, classifications and categories, criteria, methodology
• Knowledge of the universals and abstractions in a field - principles and generalizations, theories and structures
• Questions like: What are the main benefits of implementing Big Data and data analytics methods for organisation?
ComprehensionDemonstrate understanding of facts and ideas by organizing, comparing, translating, interpreting, describing, and stating the main ideas
• Translation, Interpretation, Extrapolation
• Questions like: Compare the business and operational models of private clouds and hybrid clouds.
ApplicationUsing new knowledge. Solve problems in new situations by applying acquired knowledge, facts, techniques and rules in a different way
• Questions like: What data analytics methods should be applied for specific data types analysis or for specific business processes and activities?
Which Big Data services architecture is best suited for medium size research organisation or company, and why??
AnalysisExamine and break information into parts by identifying motives or causes. Make inferences and find evidence to support generalizations
• Analysis of elements, relationships, organizational principles
• Questions like: What data analytics methods and services are required to support typical business processes of a web trading company? Give suggestions how these
services can be implemented with the selected data analytics platform, including on-premises or outsourced to cloud. Provide references to support your statements.
SynthesisCompile information together in a different way by combining elements in a new pattern or proposing alternative solutions
• Production of a unique communication, a plan, or proposed set of operations, derivation of a set of abstract relations
• Questions like: Describe the main steps and tasks for implementing data analytics and data management services for an example company or research organisation?
What services and data analytics can be moved to clouds and which will remain at the enterprise premises and run by company’s personnel?
EvaluationPresent and defend opinions by making judgments about information, validity of ideas or quality of work based on a set of criteria
• Judgments in terms of internal evidence or external criteria
• Questions like: Do you think that implementing Agile Data Driven Enterprise model creates benefits for enterprises, short term and long term?
EDISON 2018 Slide DeckData Science Professional Education and
Training53
Data Science Model Curriculum (MC-DS)
Data Science Model Curriculum includes
• Learning Outcomes (LO) definition based on CF-DS– LOs are defined for CF-DS competence groups and for all
enumerated competences
– Knowledge levels: Familiarity, Usage, Assessment (based in Bloom’s Taxonomy)
• LOs mapping to Learning Units (LU)– LUs are based on CCS(2012) and universities best practices
– Data Science university programmes and courses inventory (interactive)http://edison-project.eu/university-programs-list
• LU/course relevance: Mandatory Tier 1, Tier 2, Elective, Prerequisite
• Learning methods and learning models (in progress)
EDISON 2018 Slide DeckData Science Professional Education and
Training54
Knowledge levels for Learning Outcomes
(defined based on Bloom’s Taxonomy)
Level Action Verbs
Familiarity Choose, Classify, Collect, Compare, Configure, Contrast, Define, Demonstrate, Describe, Execute, Explain, Find, Identify, Illustrate, Label, List, Match, Name, Omit, Operate, Outline, Recall, Rephrase, Show, Summarize, Tell, Translate
Usage Apply, Analyze, Build, Construct, Develop, Examine, Experiment with, Identify, Infer, Inspect, Model, Motivate, Organize, Select, Simplify, Solve, Survey, Test for, Visualize
Assessment Adapt, Assess, Change, Combine, Compile, Compose, Conclude, Criticize, Create, Decide, Deduct, Defend, Design, Discuss, Determine, Disprove, Evaluate, Imagine, Improve, Influence, Invent, Judge, Justify, Optimize, Plan, Predict, Prioritize, Prove, Rate, Recommend, Solve
EDISON 2018 Slide DeckData Science Professional Education and
Training55
Data Science Data Analytics (KAG1 – DSDA)
related courses
• KA01.01 (DSDA/SMDA) Statistical methods, including Descriptive statistics,
exploratory data analysis (EDA) focused on discovering new features in the data,
and confirmatory data analysis (CDA) dealing with validating formulated
hypotheses;
• KA01.02 (DSDA/ML) Machine learning and related methods for information
search, image recognition, decision support, classification;
• KA01.03 (DSDA/DM) Data mining is a particular data analysis technique that
focuses on modelling and knowledge discovery for predictive rather than purely
descriptive purposes;
• KA01.04 (DSDA/TDM) Text analytics applies statistical, linguistic, and structural
techniques to extract and classify information from textual sources, a species of
unstructured data;
• KA01.05 (DSDA/PA) Predictive analytics focuses on application of statistical
models for predictive forecasting or classification;
• KA01.06 (DSDA/MODSIM) Computational modelling, simulation and optimisation.
EDISON 2018 Slide DeckData Science Professional Education and
Training56
Data Science Engineering (KAG2-DSENG)
• KA02.01 (DSENG/BDI) Big Data infrastructure and technologies, including NOSQL databased, platforms for Big Data deployment and technologies for large-scale storage;
• KA02.02 (DSENG/DSIAPP) Infrastructure and platforms for Data Science applications, including typical frameworks such as Spark and Hadoop, data processing models and consideration of common data inputs at scale;
• KA02.03 (DSENG/CCT) Cloud Computing technologies for Big Data and Data Analytics;
• KA02.04 (DSENG/SEC) Data and Applications security, accountability, certification, and compliance;
• KA02.05 (DSENG/BDSE) Big Data systems organization and engineering, including approached to big data analysis and common MapReduce algorithms;
• KA02.06 (DSENG/DSAPPD) Data Science (Big Data) application design, including languages for big data (Python, R), tools and models for data presentation and visualization;
• KA02.07 (DSENG/IS) Information Systems, to support data-driven decision making, with focus on data warehouse and data centers.
EDISON 2018 Slide DeckData Science Professional Education and
Training57
KAG3-DSDM: Data Management group: data curation,
preservation and data infrastructure
DM-BoK version 2 “Guide for
performing data management”
– 11 Knowledge Areas
(1) Data Governance
(2) Data Architecture
(3) Data Modelling and Design
(4) Data Storage and Operations
(5) Data Security
(6) Data Integration and
Interoperability
(7) Documents and Content
(8) Reference and Master Data
(9) Data Warehousing and Business
Intelligence
(10) Metadata
(11) Data Quality
EDISON 2018 Slide DeckData Science Professional Education and
Training
Other Knowledge Areas motivated by
RDA, European Open Data initiatives,
European Open Data Cloud
(12) PID, metadata, data registries
(13) Data Management Plan
(14) Open Science, Open Data, Open
Access, ORCID
(15) Responsible data use
58
• Highlighted in red: Considered (Research) Data Management literacy (minimum required knowledge)
Outcome Based Educations and Training Model
From Competences and DSP
Profiles
to Learning Outcomes (LO)
and
to Knowledge Unites (KU) and
Learning Units (LU)
• EDSF allow for customized
educational courses and training
modules design
EDISON 2018 Slide DeckData Science Professional Education and
Training59
EDSF Data Model and API
• EDSF API provides access to all EDSF functionality
EDISON 2018 Slide DeckData Science Professional Education and
Training60
DSP04 – Data Scientist MC structure
EDISON 2018 Slide DeckData Science Professional Education and
Training61
DSP10 – Data Steward MC structure
EDISON 2018 Slide DeckData Science Professional Education and
Training62
DSP04 Data Scientist – Required practical skills
and Hands-on labs
Data Science curriculum should include the following elements to achieve necessary skills Type B:
• Python (or R) and corresponding data analytics libraries
• NoSQL and SQL Databases (Hbase, MongoDB, Cassandra, Redis, Accumulo, MS SQL, My SQL, PostgreSQL, etc.)
• Big Data Analytics platforms (Hadoop, Spark, Data Lakes, others)
• Real time and streaming analytics systems (Flume, Kafka, Storm)
• Kaggle competition, resources and community platform, including rich data sets, forum and computing resources
• Visualisation software (D3.js, Processing, Tableau, Julia, Raphael, etc.)
• Web API management and web scrapping
• Git versioning system as a general platform for software development
• Development Frameworks: Python, Java or C/C++, AJAX (Asynchronous Javascript and XML), D3.js (Data-Driven Documents), jQuery, others
• Cloud based Big Data and data analytics platforms and services, including large scale storage systems
– Essential for workplace alignment
EDISON 2018 Slide DeckData Science Professional Education and
Training63
Hybrid Data Science Education Environment
(DSEE)
Hybrid DSEE and VDLabs extends regular compute and
storage resources with cloud based
• Microsoft Azure Data Lakes Analytics, Power BI, HDInsight
Hadoop as a Service, others
• AWS Elastic MapReduce (EMR), QuickSight, Kinesis and
wide collection of open datasets
• IBM Data Science Experience, Data Labs, Watson Analytics
• Google Cloud Platform (GCP)
EDISON 2018 Slide DeckData Science Professional Education and
Training64
Competences assessment and Team building
• Data Science competences assessment
(benchmarking)
• Data Science team building
EDISON 2018 Slide DeckData Science Professional Education and
Training65
Individual Competences Benchmarking
Individual Education/Training Path
based on Competence benchmarking
• Red polygon indicates the chosen
professional profile: Data Scientist
(general)
• Green polygon indicates the candidate or
practitioner competences/skills profile
• Insufficient competences (gaps) are
highlighted in red– DSDA01 – DSDA06 Data Science Analytics
– DSRM01 – DSRM05 Data Science Research
Methods
• Can be use for team skills match marking and
organisational skills management
[ref] For DSP Profiles definition and for enumerated
competences refer to EDSF documents CF-DS and DSP
Profiles.
EDISON 2018 Slide DeckData Science Professional Education and
Training66
Competence/Knowledge gap -> Suggested
LUs/courses
Recommended courses
DSDA
• Statistical Methods
• Machine Learning
• Predictive and
Quantitative analytics
• Graph Data Analysis
• Data preparation and
preprocessing
• Performance Analysis
Recommended courses
DSRM
• Research Methods and
Project Management
EDISON 2018 Slide DeckData Science Professional Education and
Training67
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
DSDA01DSDA02
DSDA03
DSDA04
DSDA05
DSDA06
DSENG01
DSENG02
DSENG03
DSENG04
DSENG05
DSENG06
DSDM01
DSDM02
DSDM03DSDM04
DSDM05
DSDM06
DSRM01
DSRM02
DSRM03
DSRM04
DSRM05
DSRM06
DSBPM01
DSBPM02
DSBPM03
DSBPM04
DSBPM05
DSBPM06
GAPS FOR DSP04
Mapping to career path:
Mastery Levels against Competences Relevance
Mastery levels defined using workplace terminology that can be easy mapped to mastery levels defined in MC-DS:
• A - Awareness 1) Understand Terminology
2) Understand Principles
3) Apply principles
4) Understand Methods
• U - Use/Application 5) Apply basics
6) Supervised use
7) Unsupervised Use
• P - Professional/Expert 8) Development of applications using wide range of technologies
9) Supervise project development, team of professionals
EDISON 2018 Slide DeckData Science Professional Education and
Training68
User/Professional level:
Measurable competences and
knowledge
Professional/Expert level:
Peer-reviewed competences
and skills
Building a Data Science Team
EDISON 2018 Slide DeckData Science Professional Education and
Training69
Data Science or Data Management Group/Department
• (Managing) Data Science Architect (1)
• Data Scientist (1), Data Analyst (1)
• Data Science Application programmer (2)
• Data Infrastructure/facilities administrator/operator: storage, cloud,
computing (1)
• Data stewards, curators, archivists (3-5)
Estimated: Group of 10-12 data specialists for research institution of 200-
300 research staff.
Growing role and demand for Data Stewards and data stewardship
EDISON 2018 Slide DeckData Science Professional Education and
Training70
Data Science or Data Management Group/Department:
Organisational structure and staffing - EXAMPLE
>> Reporting to CDO/CTO/CEO• Providing cross-organizational
services
Data Stewardship in Research and FAIR
Principles
• FAIR Initiative by Dutch Techcentre for Life Science
(DTLS) – Prof. Barend Mons– Supported by Germany, France, Spain, UK, USA
– Part of Horizon 2020 Programme
• FAIR Principles for research data:
Findable – Accessible – Interoperable - Reusable
• Data Stewards as a key bridging role between Data
Scientists as (hard)core data experts and scientific
domain researchers (HLEG EOSC report)
• Current definition of the Data Steward (part of Data
Science Professional profiles)– Data Steward is a data handling and management
professional whose responsibilities include planning,
implementing and managing (research) data input, storage,
search, and presentation.
– Data Steward creates data model for domain specific data,
support and advice domain scientists/ researchers during the
whole research cycle and data management lifecycle.
EDISON 2018 Slide DeckData Science Professional Education and
Training71
HLEG report on
European Open
Science Cloud
(October 2016)
Data Management and Governance (DMG) and
Research Data Management (RDM)
• RDM curricula example
• DAMA Data Management Body of Knowledge
(DMBOK)
EDISON 2018 Slide DeckData Science Professional Education and
Training72
Research Data Management Model Curriculum –
Part of the EDISON Data Literacy Training
A. Use cases for data management and stewardship
• Preserving the Scientific Record
B. Data Management elements (organisational and individual)
• Goals and motivation for managing your data
• Data formats
• Creating documentation and metadata, metadata for discovery
• Using data portals and metadata registries
• Tracking Data Usage
• Handling sensitive data
• Backing up your data
• Data Management Plan (DMP) - to be a part of hands on session
C. Responsible Data Use Section (Citation, Copyright, Data Restrictions)
D. Open Science and Open Data (Definition, Standards, Open Data use and reuse, open government data)
• Research data and open access
• Repository and self- archiving services
• ORCID identifier for data
• Stakeholders and roles: engineer, librarian, researcher
• Open Data services: ORCID.org, Altmetric Doughnut, Zenodo
E. Hands on:
a) Data Management Plan design
b) Metadata and tools
c) Selection of licenses for open data and contents (e.g. Creative Common and Open Database)
EDISON 2018 Slide DeckData Science Professional Education and
Training73
Collaboration with the Research Data Alliance (RDA) on developing model curriculum on Research Data Literacy: • Modular, Customisable,
Localised, Open Access • Supported by the network of
trainers via resource swap board
KAG3-DSDM: Data Management group: data curation,
preservation and data infrastructure
DM-BoK version 2 “Guide for
performing data management”
– 11 Knowledge Areas
(1) Data Governance
(2) Data Architecture
(3) Data Modelling and Design
(4) Data Storage and Operations
(5) Data Security
(6) Data Integration and
Interoperability
(7) Documents and Content
(8) Reference and Master Data
(9) Data Warehousing and Business
Intelligence
(10) Metadata
(11) Data Quality
EDISON 2018 Slide DeckData Science Professional Education and
Training
Other Knowledge Areas motivated by
RDA, European Open Data initiatives,
European Open Data Cloud
(12) PID, metadata, data registries
(13) Data Management Plan
(14) Open Science, Open Data, Open
Access, ORCID
(15) Responsible data use
74
• Highlighted in red: Considered (Research) Data Management literacy (minimum required knowledge)
DMBOK: Data Governance
and Stewardship
Scope of a Data Governance
Programme
• Strategy
• Policy
• Standards and quality
• Oversight
• Compliance
• Issue management
• Data management projects
• Data asset valuation
EDISON 2018 Slide DeckData Science Professional Education and
Training75
DMBOK: Data Management Principles
• Data is an asset with unique properties
• The value of data can and should be expressed in economic terms
• Managing data means managing the quality of data
• It takes Metadata to manage data
• It takes planning to manage data
• Data management requirements must drive Information Technology decisions
• Data management is cross-functional; it requires a range of skills and expertise
• Data management requires an enterprise perspective
• Data management must account for a range of perspectives
• Data management is lifecycle management
• Different types of data have different lifecycle characteristics
• Managing data includes managing the risks associated with data
• Effective data management requires leadership commitment
EDISON 2018 Slide DeckData Science Professional Education and
Training76
DMBOK: Data Governance Organisation Parts
• Separation or
governance
responsibilities
• Multi-layer
• CDO
• CIO
• Councils
EDISON 2018 Slide DeckData Science Professional Education and
Training77
Data Stewardship
• Creating and managing core Metadata: Definition and management of
business terminology, valid data values, and other critical Metadata.
• Documenting rules and standards: Definition/documentation of business rules,
data standards, and data quality rules.
– High quality data are often formulated in terms of rules rooted in the business
processes that create or consume data.
– Stewards help surface these rules and ensure their consistent use.
• Managing data quality issues: Stewards are often involved with the
identification and resolution of data related issues or in facilitating the process of
resolution.
• Executing operational data governance activities: Stewards are responsible
for ensuring that, day-today and project-by-project, data governance policies and
initiatives are adhered to. They should influence decisions to ensure that data is
managed in ways that support the overall goals of the organization.
“Best Data Steward is not made but found” DMBOK1 (2009)
EDISON 2018 Slide DeckData Science Professional Education and
Training78
Roadmap recommendations: Data Science
Education and Training for Europe
A. Policy recommendations for the EC and Member States
• R1. Critical skills management and training
• R2. Gender balance and multi-cultural environment
• R3. Common data literacy
• R4. Data driven technology and education divide.
• R5. European studies on demand and role of Data Science and Analytics skills
• R6. Include the above actions in the European Digital Single Market Scoreboard
B. Recommendations to universities and professional training organisations
• R7. EDSF adoption by universities
• R8. Addressing Data Science professional and workplace skills in university curricula
• R9. Multiple delivery form
• R10. Sharing experience, courses and instructors
• R11. Supporting technical infrastructure for Data Science and data related education and professional training
C. Recommendations for Research (e) Infrastructures
• R12. Critical skills management in Research (e) Infrastructures
D. Required support and contribution to the standardisation bodies
• R13. The following are essential measures to achieve EDSF support existing and new standards.
EDISON 2018 Slide DeckData Science Professional Education and
Training79
EDISON project Deliverable D4.4 Content
• EU strategy for European Digital Single Market and European
Open Science Cloud
• Overview of Recent Studies on Data Science and Analytics skills
demand
• EDISON Data Science Framework (EDSF) and Educational
Model
• EDISON Data Science Framework (EDSF) and Educational
Model
• Priority, Timeline and ongoing activities
Ongoing activities and developmentshttps://github.com/EDISONcommunity/EDSF
EDISON 2018 Slide DeckData Science Professional Education and
TrainingSlide_80
• EDISON Data Science Framework Release 3 – Published December 2018
– Call for contribution – EDSF use cases and guidelines - Deadline 30 March 2019
– Call for sponsorship
• Industry digitalisation projects and data literacy skills training– MATES project funded by EU ERASMUS Programme - EU Maritime industry digital
transformation, data skills development and Data + Ocean literacy
– Skills for SME: Data Science, IoT, Cybersecurity: Prepare to Data Economy and
Industry 4.0
• DARE project by APEC (Asia Pacific Economic Cooperation)
– Continuing cooperation for Asia Pacific region
– Recommended Data Science and Analytics Competences published August 2017 -
https://www.apec.org/Press/Features/2017/0620_DSA
EDISON Initiative Online Presence
• EDSF github project - https://github.com/EDISONcommunity/EDSF
– Component documents CF-DS, DS-BoK, MC-DS, DSPP
• EDISON Community work area and discussions -
https://github.com/EDISONcommunity/EDSF/wiki/EDSFhome
• Mailing list - [email protected]
• EDISON project website (still active) http://edison-project.eu/
EDISON 2018 Slide DeckData Science Professional Education and
Training81
Links to EDISON Resources
• EDISON Data Science Framework Release 3 (EDSF)
https://github.com/EDISONcommunity/EDSF
Component EDSF documents
CF-DS – Data Science Competence Framework
https://github.com/EDISONcommunity/EDSF/blob/master/EDISON_CF-DS-release2-v08.pdf
DS-BoK – Data Science Body of Knowledge
https://github.com/EDISONcommunity/EDSF/blob/master/EDISON_DS-BoK-release2-v04.pdf
MC-DS – Data Science Model Curriculum
https://github.com/EDISONcommunity/EDSF/blob/master/EDISON_MC-DS-release2-v03.pdf
DSPP – Data Science Professional profiles
https://github.com/EDISONcommunity/EDSF/blob/master/EDISON_DSPP-release2-v05.pdf
EDISON 2018 Slide DeckData Science Professional Education and
Training82
Other related links
• Amsterdam School of Data Science
– https://www.schoolofdatascience.amsterdam/
– https://www.schoolofdatascience.amsterdam/education/
• Research Data Alliance interest Group on Education and Training on Handling of
Research Data (IG-ETHRD)
– https://www.rd-alliance.org/groups/education-and-training-handling-research-data.html
EDISON 2018 Slide DeckData Science Professional Education and
Training83
This work is licensed under the Creative Commons Attribution 4.0 International License.
To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
Additional materials
EDISON 2018 Slide DeckData Science Professional Education and
Training84
EDSF Recognition, Endorsement and Implementation
• DARE (Data Analytics Rising Employment) project by APEC (Asia Pacific Economic Cooperation)
– DARE project Advisory Council meeting 4-5 May 2017, Singapore
• PcW and BHEF Report “Investing in America’s data science and analytics talent” April 2017
– Quotes EDSF and Amsterdam School of Data Science
• Dutch Ministry of Education recommended EDSF as a basis for university curricula on Data Science
– Workshop “Be Prepared for Big Data in the Cloud: Dutch Initiatives for personalized medicine and health research & toward a national action programmefor data science training”, Amsterdam 28 June 2016
– Currently working with Dutch Gov on re-skilling IT/data workers for DSA competences
• European Champion Universities network
– 1st Conference (13-14 July, UK)
– 2nd Conference (14-15 March, Madrid, Spain)
– 3rd Conference 19-20 June 2017, Warsaw
EDISON 2018 Slide DeckData Science Professional Education and
Training85
e-CFv3.0 Internal Structure: Refactoring for CF-DS
• s
EDISON 2018 Slide DeckData Science Professional Education and
Training86
• 4 Dimensions– Competence Areas
– Competences
– Proficiency levels
– Skills and Knowledge
• 5 Competence Areas defined by ICT Business Process stages– Plan
– Build
– Deploy
– Run
– Manage
-> Refactor to Scientific Research (or Scientific Data) Lifecycle
– See example of RI manager at IG-ETRD wiki and meeting
• Each competence has 5 proficiency levels – Ranging from technical to engineering to
management to strategist/expert level
• Knowledge and skills property are defined for/by each competence and proficiency level (not unique)
CWA 16458 (2012): European ICT Professional
Profiles
EDISON 2018 Slide DeckData Science Professional Education and
Training87
• The CWA defines 23 main ICT profiles the most widely used by organisations
CWA Professional Profiles and Organisational
Workflow
EDISON 2018 Slide DeckData Science Professional Education and
Training88